© 2009 Poultry Science Association, Inc. Evaluation of plating media for recovering 1 Salmonella from thermally treated egg albumen J. B. Gurtler2 Food Safety Intervention Technologies Research Unit, USDA, Agricultural Research Service, Eastern Regional Research Center, 600 E. Mermaid Lane, Wyndmoor, PA 19038-8551 Primary Audience: Quality Assurance Personnel, Researchers, Egg Processing Plant Managers, Regulators SUMMARY Salmonella spp. present in pasteurized egg albumen are often difficult to recover by direct plating because of thermal injury and the presence of innate iron binding and other antimicrobi- als in egg white. The literature has reported a multiplicity of selective and nonselective media used to recover heat-injured Salmonella, to measure the proportion of injured cells, or both. This study compared the proficiency of selective and nonselective plating media for support- ing colony development or for assessing bacterial injury of heat-stressed Salmonella from egg albumen. A 6-strain composite of Salmonella was added to albumen (pH 9.0), heated at 53.3°C for 3.1 min, and plated on 26 nonselective and 22 selective media. Recovery of heat-injured sal- monellae varied little (≤0.52 log cfu/mL) among the nonselective media tryptic soy agar, plate count agar, dextrose tryptone agar, or brain heart infusion agar, regardless of the manufacturer. Selective media that were optimal for recovery of salmonellae from albumen included 3 bril- liant green agars, Levine eosin methylene blue agar, and bismuth sulfite agar, which recovered more cells (P ≤ 0.05) than selenite-cystine, tetrathionate, xylose-lysine-tergitol-4 , xylose lysine deoxycholate, or Rappaport-Vassiliadis agars. The results of this study may assist in choosing selective or nonselective media to maximize the recovery of Salmonella from thermally treated albumen. Key words: egg albumen, Salmonella , medium, pasteurization, recovery 2009 J. Appl. Poult. Res. 18 :297–309 doi: 10.3382/japr.2008-00109 DESCRIPTION OF PROBLEM through the USDA Food Safety and Inspection Service (FSIS) by mandating minimum thermal The USDA National Agricultural Statistics pasteurization requirements for specified egg Service reported that US egg companies pro- products. These standards were based on data duced 2.61 billion pounds of liquid egg in 2007 for the inactivation of Salmonella in liquid egg from 24 billion eggs [1]. The Egg Products products acquired before 1970 and are currently Inspection Act of 1970 regulates egg products being reevaluated in light of recent risk assess- 1 Mention of trade names or commercial products in this article is solely for the purpose of providing specific information and does not imply recommendation or endorsement by the USDA. 2 Corresponding author: [email protected] 298 JAPR: Research Report ments, which take into account changes in in- fied M9 minimal salts broth, solidified tryptic dustrial practices such as in-line egg processing soy broth [13], solidified Dey-Engley neutral- and variations in egg product formulations [2, izing medium [14], plate count agar [15], and 3]. Regulations may be amended in accordance tryptic soy agar [16]. Selective plating media with Hazard Analysis and Critical Control Point that have been used to recover Salmonella in and Sanitation Standard Operating Procedure various states of injury by direct plating include compliance guidelines so as to incorporate le- tryptic soy agar + 0.6% yeast extract with 5% thality-based performance standards [4]. sodium chloride [17], bismuth sulfite agar [18], Pasteurized egg products have not been rec- xylose-lysine-tergitol-4 agar [18, 19], solidi- ognized as a source of human illness in the Unit- fied selenite broth, solidified tetrathionate broth ed States since the Egg Products Inspection Act [20], Levine eosin methylene blue agar with 2% of 1970 was issued. However, the USDA FSIS sodium chloride [21], MacConkey sorbitol agar has concluded that current process standards [22], xylose lysine deoxycholate agar, brilliant are not completely effective in eliminating Sal- green agar [10], selenite-cystine agar, Rappa- monella from all egg products, and predictive port-Vassiliadis agar [23], Hektoen enteric agar, models estimate that up to 5,500 people per year Rambach Plus agar, and modified lysine iron may contract salmonellosis from pasteurized agar [24]. egg products [2, 5]. The FSIS reported that dur- Reactive oxygen species (ROS) that are pres- ing routine USDA surveillance, Salmonella was ent in plating media impede the resuscitation of recovered from 0.47% of pasteurized egg prod- injured bacteria. Hydroxyl radicals and other ucts between the years of 1998 and 2003 [4]. ROS produced during autoclave-sterilization of The USDA FSIS risk assessment for Salmonella media, or shortly thereafter, are highly toxic to in shell eggs reported that current minimum bacterial DNA, membrane lipids, and cellular pasteurization standards impart 5.9, 5.5, and 3.3 proteins [25]; thus, various oxygen tension-re- log10 reductions of Salmonella in liquid whole ducing compounds have been added to selective egg, liquid egg yolk, and liquid egg white (albu- and nonselective media to enhance the recovery men), respectively [2]. They estimated that if all of injured bacteria [26]. liquid egg whites in the United States were pas- This study was designed to assess the rela- teurized for a 6-log10 reduction of Salmonella, tive proficiency of 22 selective and 26 nonselec- the annual number of human illnesses from pas- tive plating media in colony development of a teurized egg albumen products could be reduced 6-strain composite of Salmonella from thermally from 2,600 to 270 per year. treated albumen to select media that will maxi- Problems relating to recovering Salmonella mize the recovery of heat-injured Salmonella or from egg products containing yolk have been indicate the level of injury for future egg-pas- reported [6–8], although recovery of Salmonella teurization studies. An investigation of selec- from albumen has proven most difficult [9, 10]. tive compounds (NaCl, KMnO4, EDTA, sodium Innate albumen antimicrobials (e.g., ovotrans- deoxycholate, and calcium hypochlorite) added ferrin, lysozyme, avidin, etc.) [3, 11] may com- to tryptic soy agar was also undertaken to de- pound injury that Salmonella incurs during termine their inhibiting or resuscitating effects pasteurization, complicating bacterial recovery on the Salmonella composite and to assess their and enumeration. Studies that are performed to potential for use in future experiments. assess the presence and levels of Salmonella in pasteurized liquid egg products by direct plat- MATERIALS AND METHODS ing require the use of media with a high level of Bacterial Strains and Preparation sensitivity that will maximize resuscitation and of Cultures for Heat Treatment enumeration of thermally injured bacteria. Non- selective plating media that have been used to Six strains [27] of Salmonella enterica were recover Salmonella in various states of injury by examined for their response to thermal stress in direct plating include brain heart infusion agar liquid egg white and recovery on 26 nonselec- [9, 12], solidified buffered peptone water, solidi- tive and 22 selective media. Cells were grown GurTLer: MeDIA FOr reCOVerING SALMONELLA 299 for 24 h in a nutritive broth and were concen- ionized water, and opened aseptically, and then trated 10-fold [28]. 0.7 mL of the heated albumen was removed with a pipette and added to 6.3 mL of 0.1% peptone Preparation of Egg Albumen solution. Serial dilutions of thermally treated al- and Recovery Media bumen were spiral plated [34] in duplicate on the surface of nonselective or selective agar. All Albumen was prepared [29] and inoculated plates were incubated at 37°C for 48 h and colo- with a 6-strain composite of salmonellae to 7.74 nies were counted [35]. The pH and oxidation- log cells/mL. All media (Tables 1 and 2) were reduction potential (ORP) of each medium and purchased in dehydrated form and prepared ac- albumen sample were measured [36] in dupli- cording to the manufacturers’ specifications [30] cate. unless noted otherwise in the References and Notes. Statistical Analysis Thermal Treatment of Inoculated Albumen Three replicate trials were conducted, and in Thermal-Death-Time Disks duplicate samples were averaged before statis- tical analysis. Data were analyzed by ANOVA Inoculated albumen was thermally treated by using SAS software version 9.1 [37]. Significant means of aluminum thermal-death-time (TDT) differences (P ≤ 0.05) in the recovery of Salmo- disks, which have been described by Jin et al. nella within selective or nonselective media [31]. Sterile TDT disks were loaded with 0.9 mL were determined by LSD. of inoculated egg white, hermetically sealed, and immersed in a recirculating water bath [32] at 53.3°C for 3.1 min. This treatment was opti- RESULTS AND DISCUSSION mized in preliminary tests (data not shown) to Heating and Cooling Times effect an approximate 2-log reduction in popula- tion and to induce thermally mediated bacterial The heating time needed to increase samples injury to cells in the surviving population. Tem- from 21°C to within 0.5°C of the target tem- peratures were monitored with type K thermo- perature (53.3°C) was 40 s, and samples were couples and electronically recorded [33]. One held for an additional 3.1 min. The cooling times thermocouple was placed directly in the center needed to lower the temperature from 53.3°C to of the water bath to measure ambient water tem- 15.0 and 0°C were 12 and 55 s, respectively. perature. A second thermocouple was affixed inside a TDT disk through a hole drilled in the Results of pH and ORP Measurements bottom of the disk, sealed with a waterproof ad- hesive, and filled with 0.9 mL of uninoculated After inoculation, the average pH of albumen albumen to monitor the heating time and treat- was 9.0 and the average ORP measurement was ment temperature for each experiment. Heating 114 mV.